SmarterSig Improve Your Betting Conference
2012
2 Day SmarterSig Improve Your Betting
Seminar
Date Saturday 8th & Sunday 9th September 9am to 5pm
Location Ramada Encore Hotel Ramada Encore Hotel)
3 mins walk from Railway station in central
Leicester.
Cost £199 for the whole weekend not including room, lunch
and dinner (Paypal button at foot of page)
Schedule Overview
·
Day 1 AM Introduction to betting data modelling with R (no
previous experience assumed) Alun Owen
·
Day 1 PM Problems with Data Analysis and Possible Solutions Steve
Tilley
·
Day 1 PM Hurdles in betting, can you really win Steve Tilley
Informal dinner and drinks in the evening and informal betting
chat
·
Day 2 AM Inroduction to data modelling
with R (follows day 1 & competition time) Alun
Owen
·
Day 2 PM Betfair API Programming John
Jackson
·
Day2 PM Money Management John Jackson
·
Day 2 PM Ratings and Oddslines - Simon Rowlands (Head of Timeform
R&D and Betfair Blogger)
Introduction to betting data modeling with R – Alun
Owen
·
R is an extremely powerful and flexible tool for
analysing data and developing and implementing statistical models.
·
It is free to use and has a huge community of users
many of whom submit ready-made R "packages" to a central library that
can be used to fit a wide variety of models to data.
·
We will begin with some of the basics such as how
to use R to examine your sports data, using summary statistics, tables and
plots etc.
·
But we will soon move on to look at how to fit
models to your data. Steve Tilley will be looking at Linear Regression models
in the afternoon of Day 1and so we will only mention that model briefly.
·
The main business of the morning session in Day 1
is to allow delegate to understand what a Logistic Regression model is and why
they are so popular in sports modelling,
but more importantly how to fit these models using R.
·
We will also look at how to assess whether our
model is any good and how to compare several models to decide which might be
the "best" one to adopt.
·
The morning session of Day 1 will be restricted to
relatively simple and clean data sets so that the focus is on the modelling
process.
·
The morning session of Day 2 we will move to data
sets that are more like those you are likely to use in practice and will
culminate in delegates (in teams) attempting to develop and deploy their own
models to real data as part of a competition.
·
We will assume no previous knowledge of R, although
a simple introductory tutorial will be supplied prior to the conference. so that delegates should to be up and running with R right
from the off!
·
We will keep any mathematical theory in the
background as much as possible, so that the implementation of the models is
kept as the main focus. So you won't need a degree in mathematics but we will
assume a level of mathematical competence equivalent to a good grade (A or B)
at 'O' level (if you are the same age as Alun) or
GCSE (if you are not as ancient as Alun) in
mathematics.
Problems with data analysis and possible solutions – Steve Tilley
·
The starting point is why is it that it
so difficult with all the available sports data and software to derive profitable
methods of betting from this data?
·
We will look at different types of data
source and the advantages and problems that they bring.
·
How much data should you have? Can you
ever have too much data?
·
We will look at what you should do before
you even start looking at the data.
·
We will look at some simple tests you
should employ before you start analysing the data.
·
We will consider different strategies
of data analysis and their difficulties.
·
We will consider different types of
variable and how they can be encoded.
·
Data handling software and their
applications.
·
What are you trying to measure? What
values are you trying to optimise?
·
We look at one of the simplest of
statistical models, linear regression, and the pitfalls that can lead to.
·
We will consider some “black box”
techniques and compare them to ordinary statistics.
·
How can you tell if a method that
worked in the past will work in the future?
·
Why Monte Carlo methods are useful for
gamblers.
Hurdles in
betting, can you really win – Steve Tilley
·
Know thyself. A quick question and
answer session so we can see where we all fit in to the spectrum of gamblers.
·
My definitions of a recreational
gambler, a semi pro gambler and a pro gambler. Which are you and more
importantly which do you want to be?
·
What are your aims and objectives? What
are the constraints that bind you?
·
Record keeping.
·
Setting up a betting bank, deciding on
stakes.
·
Losing runs and how to deal with them
·
Winning runs and how to deal with them.
·
Boredom, the biggest enemy.
·
Preparations that will save you money.
·
Why “fun bets” generally aren’t.
·
How much can you lose and still keep
going? Ways of dealing with those real crunching losses.
·
Why big wins can be dangerous.
·
Dealing with your betting firms.
·
Dealing with other gamblers.
·
Dealing with friends and family.
·
Making the whole thing more enjoyable.
Getting On - Betfair API Programming - John Jackson
·
Creating a program to interact with Betfair
from scratch
·
Code segments will be provided for the whole session
·
The importance of testing to avoid calamities
·
Saving years of programming: R and the betfaiR packages for Windows
·
Automatically logging in, maintaining communication and logging
out
·
Extracting Market Data
·
Placing and cancelling bets automatically
·
Contrasting Windows and Unix programming
·
Discussion of betting strategies if time permits
If everyone is comfortable then ...
·
Asynchronous price extraction
·
Timed operations
·
Q&A session for usage scenarios e.g. greening up; monitoring
for specific circumstances ; continuous trading; ...
<your question here>.
Ratings and Oddslines - Simon Rowlands
·
Introduction:
purpose is to give overview of ratings and to give Conference attendees ideas
for compiling their own ratings or of using existing ratings in different ways,
including as a basis for calculating odds lines
·
Limitations
of some conventional methods of producing ratings
·
Getting
started
·
Standardisation
(handicap standards; non-handicap standards; prior-rating standards; breeding
standards etc)
·
A
quick overview of pounds-per-length (derived from pounds-per-second)
·
Back-handicapping,
including using market information to create exceptions
·
Using
ratings to measure other factors in racing (draw, trainer form etc)
·
Using
ratings predictively rather than descriptively
·
Using
ratings to prompt odds lines (including brief consideration of weighted
ratings, Monte-Carlo simulation and chains)
·
Possible
future uses of ratings and odds lines
·
Question
and answer
Money
Management - John Jackson
·
Methods used by professionals for reducing the ups and downs and increasing
profit
·
Developing a long term strategy
·
Defining your goal
·
"all staking problems are forms of dynamic programming"
·
The characteristics of level staking; proportional staking and
progressive betting
·
Balancing risk and reward: The Sharpe Ratio
·
Operational hedges: insurance; camouflage; commission reduction
·
Equivalent bets; synthesising bets not offered by layers
·
The role of time in betting
·
'safe' bets and leverage
·
Using hedging in combination with long term strategy
·
Never forget the lesson of LTCM
Equipment
To get the best out
of the 2 days you will need to bring with you a windows
based laptop computer. One or two programs will be best installed pre
conference but this is a simple procedure and we will be giving instructions
and probably a video on how to do this shortly.
Overnight Stays
The best hotel deals
tend to be via online bookings. You can book a room at the Encore
hotel (negotiated rate of £65 or alternatively the Premier Inn is right next door to the train station
and only 2 minutes walk from the conference venue.
http://www.premierinn.com/en/hotel/LEIBAR/leicester-city-centre
The Ibis hotel is also only 3 minutes walk from the station and about the same to the conference venue
http://www.ibishotel.com/gb/hotel-3061-ibis-leicester-city/index.shtml
Lunch
There are numerous options for lunch within 60 seconds walk of the
hotel entrance, these include
·
Nandos
·
Coffee Republic
·
Cafe Neros
·
Subway Sandwich/coffee
·
La Tasca (Spanish)
·
Two Italian restaurants
Evening Meal
The plan around 7pm will be to eat together at the Encore hotel where I have negotiated a half price rate on food for the evening. This is of course optional but in the past these have proved
excellent networking opportunities and we will no doubt have a fun sweepstake from Wolverhampton races
Conference
Payment
To register an interest in attending please email 
It makes it a lot easier to learn, improves confidence to try things and motivation to use these tools further.
I felt the quality of speakers were really high. Dr Alun Owens introductory sessions on R were a particular highlight. To be taken through with examples of how to use the software in a meaningful way regarding betting has broken down the mystery of how to start getting into this software environment. The scripts and documentation I feel are also especially useful as the ability to edit and reverse engineer their structure will accelerate my learning of its language.
It was just good to listen to other people for once, personally, rather than on some forum; no masks, bouncing ideas, learning, thinking, knowing I'm not the only one in the boat regarding bookie restrictions, etc. Good for motivation too, seeing as though I'm relatively new to all this (2.5 years betting/trading).
As important a part of the weekend (evening dinner and drinks) as the talks if not more so. Sweepstake was an excellent idea and reminded me how good Steve's challenges can be (remember the Cotswold terrier challenge). Networking opportunities over the whole weekend were very good.
Great venue, easy to mix with people.
I felt that the topics covered were varied but well well balanced. I personally learnt much more than I had hoped for from each of the speakers and the tools and ideas which I have taken away with me will certainly Improve my Betting.
I feel that it was a good and fair idea to host the conference in a city which is located in the center of the country rather than London. Conference hotel venue was ideally located; very happy with the choice of eating options and (cheaper!) hotels nearby as they were only a short walk away.
Very much enjoyed the chance to socialise over a meal with the opportunity for a few drinks afterwards. Made it much easier to get to know people and build on basic introductions at the conference.
The variety of topics on offer ensured there was something useful for everybody there. Having worked examples for the R and API talks was also extremely useful. The football sweepstake in the evening was great fun and went a long way to breaking the ice. Overall a thoroughly enjoyable weekend and I look forward to attending the next installment!
Conference Feedback From Delegates